@eyedeekay
&zzz
+R4SAS
+RN
+StormyCloud
+T3s|4
+acetone
+dr|z3d
+hk
+orignal
+postman
+radakayot
+snex
+weko
+wodencafe
Arch
BravoOreo
Dann
FreeB
FreefallHeavens_
Irc2PGuest11045
Irc2PGuest48814
Irc2PGuest59134
Irc2PGuest60478
Irc2PGuest7448
Irc2PGuest90968
Leopold
Onn4l7h
Onn4|7h
Sleepy_
Soni
T3s|4_
Teeed
aeiou
aisle
ardu
b3t4f4c3__
bak83_
dickless
dr4wd3
enoxa
eyedeekay_bnc_
hagen_
not_bob_afk
phil
plap
poriori
profetikla
qend-irc2p
rapidash
solidx66_
u5657
uop23ip
w8rabbit
x74a6h
EKCKABATOR54
Why are there NTCP2 and SSU2? Its possible to write tcp over udp and use NTCP2 even over UDP. Or there are 2 transports, so that if one is blocked, the second remains or there are some other reasons?
zzz
it's if one is blocked, + other reasons
EKCKABATOR54
Also, do you have plans to add java i2p to this tool github.com/h-phil/i2pd-testnet-kubernetes?
zzz
SSU2 provides address detection, firewall detection, and NAT traversal
EKCKABATOR54
Thank you for the answer
zzz
the h-phil project was done by students who have probably moved on
zzz
we did make the suggestion last year, perhaps the next batch of students will carry it forward
zzz
huge 117-file PQ WIP MR is up, in case I get hit by a bus, or if orignal wants to peek at how I did something git.idk.i2p/i2p-hackers/i2p.i2p/-/merge_requests/239
zzz
looking for test results on 2.8.1-1-rc from anybody that hit the SHA256 bug, including RN
zzz
eyedeekay, let's do some 2.8.2 planning, I volunteer to do the release, maybe late this week?
zzz
woo the MR passed CI, I didn't forget any files
dr|z3d
nice, zzz
dr|z3d
the bug I referenced the other day probably wasn't a bug, as I suspected, looks like a memory management issue in the OS. gone.
zzz
good
zzz
so nobody's seen the SHA256 error except me, RN, and afkaid iirc?
dr|z3d
not seen it, your patch is rolled out to +, so hopefully nobody will.
zzz
does your entire userbase update every day? or?
dr|z3d
no. fairly regularly, though, if the logs are to be believed.
dr|z3d
some from the dev path, some from release. iirc I pushed out an update with your patch for both.
zzz
ok
eyedeekay
zzz late this week works for me, thank you for volunteering to do it I've got a bit of a backlog here, Thursday-Friday? Friday-Saturday? What's best for you?
zzz
eyedeekay, let's defer the decision on timing for a day or two, if that's ok, or if not, please propose one, I'm flexible
dr|z3d
not my call to make, but if you want to be somewhat consistent, Monday is probably "a good idea".
zzz
I still don't know why everyone's not seeing the logs, it seems like they should; my guess is some JVM or traffic quirk, so my bias is to sooner rather than later
dr|z3d
yeah, I suspect it's probably not present in java 21 or later.
dr|z3d
have you seen it on any box running >=21?
zzz
not running any on 21+
zzz
I ran it for 8 weeks on 3 routers and never saw it once, but when I put it on a different router on saturday it was tripping serveral times an hour
dr|z3d
could be a good time to get a router on 21 or later.
zzz
maybe. My boxes are on 8, 11, and 17, and I almost always build on 8, to make sure no 9+ stuff sneaks in.
dr|z3d
sure, I'm always building on 8 until we decide to bump the min version.
eyedeekay
Sounds like there are still some mysteries to solve, let's call it Monday unless the answers become clear in the meantime
zzz
I'm saying the mysteries may not be solvable and all indications are that everybody has the bug, just not everybody is getting logs, so that's why I'm leaning to this week
eyedeekay
Ok well then we can defer the decision a couple days and decide tomorrow or Wed, but I'd appreciate a 1-day buffer between decision day and release day
zzz
ok
orignal
zzz, about ML-KEM-512
orignal
seems cipher key after encaps is 768 bytes instead 800
orignal
is it right?
zzz
orignal, correct, the two directions are different sizes
orignal
it's not mentioned in the proposal
zzz
I called those keys MLKEMxxx_CT (for ciphertext) in the proposal
zzz
didn't explicitly say it but the sizes are in the charts in the proposal
zzz
or you can always double-check in FIPS-203
orignal
thanks
orignal
that's what openssl call syas
zzz
haven't heard that term
orignal
says
orignal
typo
orignal
so I called openssl function and it returned 768 instead 800
zzz
yup
orignal
got encaps/decaps pair working
zzz
nice
orignal
ready to start ratchets with 5
RN
zzz checking logs
RN
net.i2p.crypto.SHA256Generator.calculateHash(SHA256Generator.java:63) ◀━━ only mention of sha256 in logs that I can see
zzz
RN are you running 2.8.1-1-rc with the fix?
RN
yes 2.8.1-1rc
zzz
and you have any of the same errors you had before?
zzz
or similar?
zzz
the previous ones were actually CRIT
RN
not seeing any, much less log noise so I can see the "Content is not allowed in prolog." thing I haven't tracked down yet
zzz
so they would be in the 'critical logs' section on the logs page and you would have had red bubbles
RN
yeah, only i2ptunnel.I2PTunnelHTTPServer: Error connecting to HTTP server /xxx.xxx.xxx.xxx:xxxxx and a couple WARN in there
RN
think that err is something I was doing and didn't finish
RN
yes, that's something where I don't have the server running right now
zzz
ok thanks, thats what I was looking for, that the CRITs went away
RN
yep
zzz
super
RN
:)
zzz
RN btw what java version?
RN
Java version:OpenJDK BSD Porting Team 18.0.2 (OpenJDK Runtime Environment 18.0.2+9-1)
zzz
thx
RN
:)
zzz
RN, almost certainly not causing you problems, but 18 is stranded, recommend you go up to 21 or down to 17 which are LTS
RN
yeah, I do need to bump that
orignal
"With ML-KEM, there are now three sections. The first section contains the encrypted PQ public key. The second section contains the static key. "
orignal
zzz
orignal
worng statment
orignal
never mind
orignal
basically my question what static is encrypted with?
orignal
same key as encaps_key section?
orignal
the spec doesn't metnion what's used for static encrypt